Details
-
Bug
-
Resolution: Fixed
-
Blocker
-
3.0
-
Security Level: Public
-
None
Description
During commit validation of a couchdb commit, the testrunner got stuck at a rebalance. It seems that ns_server tries to move buckets, but it can't due to:
[rebalance:debug,2013-11-07T17:08:38.704,n_3@127.0.0.1:<0.990.0>:janitor_agent:do_wait_checkpoint_persisted:832]Got etmpfail waiting for checkpoint persistence. Will try again
[ns_server:info,2013-11-07T17:08:38.905,babysitter_of_n_3@127.0.0.1:<0.86.0>:ns_port_server:log:169]memcached<0.86.0>: Thu Nov 7 17:08:38.704324 CET 3: (default) Notified the timeout on checkpoint persistence for vbucket 20, cookie 0x52b8f00
In case you want to reproduce it. Just:
1. checkout and build the current master
2. in the ns_server directory run: ./cluster_run -n 4
3. in the testrunner directory run: ./testrunner -i b/resources/dev-4-nodes.ini makefile=True -t rebalance.rebalancein.RebalanceInTests.incremental_rebalance_in_with_queries,blob_generator=False,items=2000,is_dev_ddoc=False,max_verify=2000